Adaptiv Compnation Mthod Uing th Prdiction Algorithm for th Dopplr Frquncy Shift in th LEO Mobil Satllit Communication Sytm Moon-H You, Song-Pal L, and Youngyarl Han In low arth orbit (LEO) atllit communication ytm, mor vr pha ditortion du to Dopplr hift i frquntly dtctd in th rcivd ignal than in ca of gotationary arth orbit (GEO) atllit ytm or trrtrial mobil ytm. Thrfor, an timation of Dopplr hift would b on of th mot important factor to nhanc prformanc of LEO atllit communication ytm. In thi papr, a nw adaptiv Dopplr compnation chm uing location information of a ur trminal and atllit, a wll a a wighting factor for th rduction of prdiction rror i propod. Th prdiction prformanc of th propod chm i imulatd in trm of th prdiction accuracy and th cumulativ dnity function of th prdiction rror, with conidring th offt variation rang of th initial input paramtr in LEO atllit ytm. Th imulation rult howd that th propod adaptiv compnation algorithm ha th bttr prformanc accuracy than Ali mthod. From th imulation rult, it i concludd th adaptiv compnation algorithm i th mot applicabl mthod that can b applid to LEO atllit ytm of a rang of altitud btwn 1,000 km and,000 km for th gnral rror tolranc lvl, M = 50 Hz. Manucript rcivd Augut 18, 000; rvid Novmbr 9, 000. Moon-H You i with th Radio&Broadcating Tchnology Laboratory, Tajon, ETRI, Kora. (phon: +8 4 860 6456, -mail: moon@tri.r.kr) Song-Pal L i with th Radio &Broadcating Tchnology Laboratory, Tajon, ETRI, Kora. (phon: +8 4 860 60, -mail: pall@tri.r.kr) Youngyarl Han i with th Diviion of Elctrical and Computr Enginring, Hanyang Univrity, Soul, Kora. (phon: +8 90 036, -mail: yyhan@mail.hanyang.ac.kr) I. INTRODUCTION In low arth orbit (LEO) atllit communication ytm, mor vr pha ditortion i frquntly dtctd in th rcivd ignal than in ca of go-tationary arth orbit (GEO) atllit ytm or trrtrial mobil ytm [1] [3]. Svr pha ditortion i from th rult of th Dopplr frquncy hift on th rcivd ignal. Thi ditortion com from th fatr movmnt of a atllit rlativ to that of a trminal and an arth tation. Sinc th rlativ vlocity of th atllit to th urfac of th arth i vry fat, th Dopplr frquncy hift affct th carrir frqunci largly in th communication link. Th tim variation of th Dopplr frquncy hift alo bcom vry larg. It ffct on communication link mak th prformanc of th rcivr to b dgradd, whr th amount of prformanc dgradation dpnd on th tranmiion chm ud in th communication ytm. Uually, th Dopplr frquncy hift i mor harmful to th digital communication ytm mploying cohrnt dmodulator. Conquntly, thr ha bn much rarch king a mthod to compnat for th Dopplr frquncy hift [4] [7]. Rcntly, a Dopplr timation chm uing rlativ tim information wa introducd [1]. Thi mthod pr-compnat for th Dopplr frquncy hift bfor carrir rcovry. In [1], a prdiction chm for Dopplr frquncy hift ha bn dcribd by uing th rlativ tim to th rfrnc tim whn th atllit mak maximum lvation angl with th ur trminal. Thi algorithm mad th prdiction quation for Dopplr frquncy in LEO atllit communication vry impl by uing th rlativ tim concpt. But aftr th initial paramtr tting in thi algorithm, th tim diffrnc i only a 3 Moon-H You t al. ETRI Journal, Volum, Numbr 4, Dcmbr 000
paramtr applid to th prdiction quation through th tim. Thrfor, th initial input paramtr hav a larg ffct on th prdiction rror. Du to thi rror rang, th algorithm i limitd in application. In ordr to rduc prdiction rror, a nw prdiction mthod uing th information of atllit phmrid i propod in thi papr. Th cyclic movmnt of a LEO atllit i rprntd by dtrminitic formula with it priod. Thrfor, if th movmnt of th trminal i ignord, th Dopplr hift, which i in proportion to th rlativ vlocity btwn a LEO atllit and a trminal, alo ha it dtrminitic function, and can b rprntd in th tim domain [8]. Uing thi in fdr link, th Dopplr hift i aily prdictd in a fixd arth tation bcau th arth tation know th xact information about it own poition and th atllit tim-varing location. But a mobil trminal in ur link may not know it own poition on tim whn it powr turn on. In thi papr, a nw compnation mthod that i abl to timat a ur trminal poition and prdict it continuou Dopplr frquncy hift imultanouly i propod. Firt of all, th Dopplr frquncy hift i dcribd in a clod form with gographical paramtr. Nxt, th Dopplr frquncy prdiction algorithm and th adaptiv compnation mthod uing th prdiction algorithm by applying th wighting factor ar propod. Th prformanc of thi nw prdiction chm i compard to that of th known mthod [1]. In thi nw prdiction mthod, location information of a ur trminal and a atllit i ud. Thn th prdiction prformanc of th propod chm by uing LEO atllit contllation ha bn imulatd. Th imulation rult will dmontrat th nhancd timation prformanc of th propod chm. Figur 1 how th gographical configuration btwn a atllit and a ur trminal., d v ( t ) = () dt a + r arcoψ (, = (3) whr a i th radiu of th atllit orbit, r i th radiu of th arth, and ψ ( i th paration angl btwn th atllit and th ur trminal a hown in Fig.1. X x Z, z r α ψ v u X P h v X θ Ω Ω θ A θ F y Y II. DOPPLER SHIFT CHARACTERISTICS IN LEO SATELLITE COMMUNICATION SYSTEM Dopplr hift i a natural phnomnon in mobil communication and cau vr pha ditortion and prformanc dgradation pcially in LEO mobil atllit communication ytm. Th Dopplr hift charactritic in LEO atllit communication ytm i analyzd a follow. Th Dopplr frquncy hift, f D ( i proportional to th carrir frquncy f c. Th rlativ vlocity btwn a tranmittr and a rcivr, v( can b xprd a following (1). v( f D ( = fc, (1) c whr c i th vlocity of th light. Th rlativ vlocity, v(, and th ditanc btwn a tranmittr and a rcivr, (, can b rprntd by () and (3). X, Y, Z: Earth-cntrd inrtial (ECI) fram x, y, z: Earth-cntrd fixd (ECF) fram : Ditanc btwn a atllit and a ur trminal r: Radiu of th arth h: Altitud of atllit a: Radiu of th atllit orbit (= r + h) P: Sub-atllit point α : Satllit lvation angl A: Acnding nod ψ : Sparation angl btwn a atllit and a ur trminal θ S: Orbital angl of th atllit orbit θ : Tim-varying longitud angl of th ur θ F: Orbital angl of th atllit orbit in th ECF fram v r S : Satllit vlocity vctor v r : Earth rotation vlocity vctor u r : Unit vctor from th atllit to th ur trminal v: Rlativ vlocity from th atllit to th ur trminal ( = r r v v S r u ) Fig. 1. Gographical configuration of th LEO atllit and th ur trminal. By uing (1), (), and (3), th Dopplr frquncy hift and it ETRI Journal, Volum, Numbr 4, Dcmbr 000 Moon-H You t al. 33
variation rat can b aily drivd a hown by (4) and (5). f& D f D fc ar d ( = coψ (4) c dt d fc a r d ( = f D ( = co 3 ψ t dt c ( dt () f ar d + c co (. ψ (5) c ( dt Whn th ytm paramtr, uch a, th carrir frquncy and th atllit altitud ar dtrmind, ψ ( i an important factor for calculating th Dopplr hift frquncy. On of th common mthod of xpring coψ ( can b found in [9], and it i hown in (6). It i charactrizd in th arth cntrd inrtial coordinat fram by uing atllit orbit angl bad on an acnding nod, θ S( and th tim-varying longitud angl of th ur, θ (. In thi fram, θ S( and θ ( can b xprd in trm of th angular vlocity of th atllit, ω and th angular vlocity of th arth, ω, rpctivly. coψ ( = cot coθ ( coθ ( + coicot inθ ( inθ ( + iniint inθ (, (6) whr, i i th orbit inclination angl and T i th latitud of th trminal poition. Compard to th claical mthod, Ali rprntd coψ ( a (7) in anothr approach uing th tim, t m at th atllit making maximum lvation angl with th ur trminal in th arth cntrd fixd coordinat fram. In thi fram, th angular vlocity of th atllit, ω F vari with latitud du to arth rotation [1]. ( θ ( θ ( t )) coψ ( t ), coψ ( = co (7) F whr, θ F( i th orbital angl of th atllit on th bai of th atllit acnding nod in th arth cntrd fixd coordination ( = ω F t + θ F(0)). In thi papr, coψ ( i rprntd a quation (8) uing th ur trminal poition, i.., th latitud, T and th longitud, G, which ar practically mor applicabl paramtr than θ (, θ ( or ψ ( t m ), θ ( t m ) in th arth cntrd fixd coordinat fram. = cot cot co( G G ) int ( int co ψ + (8) whr T and G ar th latitud and th longitud of th atllit poition, rpctivly. It i aumd that th ur trminal can aily obtain thi information of atllit phmrid at tim t. F m m Actually, many atllit uch a Globaltar [10] and KOMP SAT [11], [1] hav th onboard GPS rcivr. In Globaltar ytm, th mobil ur trminal can dtrmin which particular atllit hould b in viw and gt th knowldg of atllit phmrid. Aftr obtaining th function of coψ ( from on of th abov quation, th Dopplr frquncy hift can b calculatd by uing (4). If th calculatd valu of coψ ( i mor accurat, th prdiction accuracy of th Dopplr frquncy hift i highr. Now th adaptiv compnation mthod i prntd uing th nw prdiction algorithm for th Dopplr frquncy hift by uing (8). III. ADAPTIVE COMPENSATION METHOD Now, it i vry clar that lction of a crtain xprion for ψ ( would dtrmin th rquird paramtr for Dopplr calculation and that th prdiction of th intantanou Dopplr frquncy would b abl to compnat th xact Dopplr frquncy. PDF (%) 0. 0.15 0.1 0.05 Propod algorithm Ali' algorithm 0-10000 -5000 0 5000 10000 Dopplr prdiction rror (Hz) Fig.. Probability dnity function of th prdiction rror of th dopplr frquncy. A mntiond abov, a prdiction chm for Dopplr frquncy hift by uing an xprion for ψ ( rprntd a (7) ha bn dcribd in [1] by Irfan Ali, whr th timation of th Dopplr frquncy i bad on th rlativ tim to th rfrnc tim whn th atllit mak maximum lvation angl with th ur trminal. Thi algorithm mad th prdiction quation for Dopplr frquncy in LEO atllit communication vry impl by uing th rlativ tim concpt. But aftr th initial paramtr tting in thi algorithm, th tim diffrnc i only a paramtr applid to th prdiction quation through th tim. Thrfor, th initial input paramtr hav a larg ffct on th prdiction rror. Th prformanc of Ali 34 Moon-H You t al. ETRI Journal, Volum, Numbr 4, Dcmbr 000
prdiction algorithm i prntd on Fig. howing th probability dnity function of th Dopplr frquncy prdiction rror. Thi rult i obtaind undr th condition that th LEO atllit i at th altitud of 780km a hown in Fig.. Th prdiction algorithm propod in [1] produc a Dopplr frquncy timation with a larg varianc of prdiction rror. Du to thi rror rang, th algorithm i limitd in application. In thi papr, to gt a bttr prformanc, nw compnation mthod i uggtd a follow. In propod chm, ky paramtr dtrmining nw prdiction algorithm for th Dopplr frquncy hift ar th latitud and longitud information of th ur trminal poition. In ordr to gt th ur trminal poition information, th ditanc function, (, btwn atllit and ur trminal ha to b found. 1. Th Algorithm Dtrmining Ditanc btwn Satllit and Ur trminal, ( In ordr to gt th ditanc, (, btwn atllit and ur trminal, an quation of th cond dgr of ( could b mad by uing (3), (4), and (5). Thn, to implify th cond drivativ of coψ ( in (5) and to rplac it with th function of (, w introduc th approximation function of th angular vlocity of th atllit in th arth cntrd fixd coordinat fram, ω F. Th quatorial componnt of ω F i dpndnt on th latitud of th ub-atllit point du to arth rotation. For LEO atllit, w can minimiz th ffct of th latitud of th ub-atllit point on th angular vlocity of th atllit and rplac th trm of th latitud of th ub-atllit point to th trm of th orbit inclination angl. Thrfor, th angular vlocity of th atllit, ω F can b approximatd by th following quation. & d θ F = θ F = ω F ω ω coi dt (9) Applying (7) and (9), w obtain th approximation of th cond drivativ of coψ ( a (10). d dt ω F ψ coψ co (10) Now, by uing (4) and (10) into (5), th following (11) of cond dgr of ( can b xprd in th clod form. By lcting th gographically corrct olution btwn two root of th quadratic (11), th ditanc ( can b obtaind. cf& ( ) D t ( ( ω fc cf D ( ) a + r ω f c = 0. (11). Th Algorithm Dtrmining th Trminal Poition, T and G W lct th rult, ( t 1 ) and ( t ), by uing th maurd Dopplr frquncy hift, f D ( t 1 ) and f D ( t ), and Dopplr hift rat, f & D ( t 1 ) and f & ( t ) at ampling intant, t D 1 and t ( > t1), rpctivly. A applying ditanc ( t 1 ) and ( t ) into (3), coψ ( t1) and coψ ( t ) can b obtaind. By applying coψ ( t1) and coψ ( t ) into (8), th imultanou quation for two paramtr, th latitud, T and th longitud, G of th trminal poition can b obtaind. By applying trigonomtric function, th trminal poition information i obtaind a hown in (1) through (14). Th lowr ubcript, 1 and hown in th quation indicat th ampling intant, t1 and t, rpctivly. whr, A n n ing cog ( A D1 A1 D ) ( AC1 A1C ) ( A B1 A1 B ) cot ( BD1 B1D ) ( BC1 B1C ) ( A1 B A B1 ) cot int = (1) int = (13) in T k ± k k1k3 = (14) k 1 = cot ( tn ) cog ( tn ), Bn = cot ( tn ) ing ( tn ) = ( tn ), Dn = coψ ( t n ) = ( A ) ( ) ( ) C1 A1C + BC1 B1C + A1 B A 1 = ( AC1 A1C )( A D1 A1 D ) + ( BC1 B1C )( BD1 B1 ) = ( A D A D ) + ( B D B D ) ( A B A ) C int k k k 1 B D 3 1 1 1 1 1 B1 3. Compnation Mthod Uing th Dopplr Frquncy Prdiction Algorithm Now, th Dopplr frquncy hift could b prdictd by uing th trminal poition paramtr, bcau tho paramtr hav bn timatd. A nw prdiction quation of th Dopplr frquncy hift, f DM 1( can b obtaind a by (15). Th ytm could continuouly prform th compnation opration againt th pha ditortion du to Dopplr hift by uing th Dopplr prdiction valu. f ( = arfc DM1 ω ω coi ω cot cot in i co in 1 int ini tant co( G G ) in( G G ) cot + ( ω in i coθ ( ) ( cot () ( () ) ( ) ) t cot co G t G int t int c a r ar + + int (15) ETRI Journal, Volum, Numbr 4, Dcmbr 000 Moon-H You t al. 35
Figur 3 how th xact Dopplr frquncy hift, f D ( and th prdictd frquncy hift, f DM 1( undr th condition that th carrir frquncy i.4ghz and givn diffrnt altitud of th atllit. In Fig. 3, it i hown that th prdictd frquncy hift, f DM 1( follow th xact Dopplr frquncy hift, f D ( vry wll, and Fig. how that th prdiction prformanc of th nw prdiction algorithm i bttr than that of th algorithm in [1] by man of th comparion btwn th varianc of th prdiction rror for th two prdiction algorithm. Dopplr frquncy (Hz) 60,000 40,000 0,000 0-0,000-40,000 h=1,000km h=1,500km xact dopplr timatd dopplr h=,000km -60,000 0 00 400 600 800 1000 100 1400 Tim (c) Fig. 3. Comparion btwn th xact Dopplr frquncy hift, f D ( and th prdictd frquncy hift, f DM1 ( at f c =.4 GHz. Conidring th ytm opration, th prdiction rror tolranc i vry important not only for tatitical raon but alo in th intantanou tim apct. Lt th intantanou rror, ( b dfind a th diffrnc btwn th xact Dopplr frquncy, f D ( and th prdictd frquncy hift, f DM 1(, and th uppr limit of th rror qual M, a th rror tolranc lvl which i gnrally adoptd a about 50Hz in ral ytm. ( = f ( f 1 ( M (16) D Figur 4 how th intantanou rror, (, of th prdictd Dopplr frquncy, f DM 1( in th ca of diffrnt altitud of atllit, in which th xact initial Dopplr frqunci, f D ( t 1 ) and f D ( t ) wr ud for th input paramtr. By obrving Fig.4, on can notic two fact, th on i that th prdiction rror ar not tolrabl during mot of th priod, and th othr i that th intantanou rror m to b dpndnt on th ditanc,, th paration angl, ψ, or th atllit lvation angl, α. In fact, thi rror i caud by th approximatd angular vlocity of th atllit, ω in (9). 4. Adaptiv Prdiction Algorithm of th Dopplr Frquncy Shift with th Wighting Factor To minimiz th rror in f DM ( ), th prdiction algorithm 1 t DM F dopplr prdiction rror (Hz) Frquncy offt (Hz) 1,400 1,00 1,000 800 600 400 00 0 0 00 400 600 800 1,000 1,00 1,400 1,600 Tim (c) h = 1,000 km h = 1,500 km h =,000 km Fig. 4. Intantanou rror of th nw prdiction algorithm. 1,00 1,000 800 600 400 00 A*in^ (alpha) Prdiction rror 0 0 15 30 45 60 75 90 Elvation angl, α (dg) Fig. 5. Prdiction rror of dopplr frquncy hift vru th atllit lvation angl. could b conidrd to th adaptiv chm by introducing th wighting factor, W. A hown on Fig. 5, th prdiction algorithm of th Dopplr frquncy hift, f DM 1(, ha rror pcially around th maximum lvation angl, and th trnd of th intantanou prdiction rror follow in-quar function of th lvation angl. To rduc th rror at highr lvation angl, a nw adaptiv prdiction algorithm i uggtd a follow. According to th approximation of th intantanou rror function mntiond abov, th wighting factor, W ( i introducd by th following quation. In th proc of th Dopplr frquncy prdiction, th atllit lvation angl i timatd by uing (18). Thrfor, th intantanou wighting factor, W ( i obtaind vry aily. W ( = Ah in α(, (17) a r ( inα ( =, (18) r( 36 Moon-H You t al. ETRI Journal, Volum, Numbr 4, Dcmbr 000
whr, A h i th normalizing factor dpnding on latitud of th atllit. Th comparion btwn th prdiction rror in (16) and th wighting factor in (17) i hown in Fig. 6. In (19), by adding th wighting factor, W ( to th timatd Dopplr frquncy, f DM 1(, th adaptiv Dopplr frquncy prdiction, f DM ( com clor to th xact Dopplr frquncy. Frquncy offt (Hz) 10,000 1,000 100 10 f h=1,000km DM ( DM1 + t h=1,500km h=,000km = f ( W ( ). (19) prdiction rror wight propod prdiction algorithm without wighting factor, propod adaptiv compnation algorithm with wighting factor and Ali prdiction algorithm. A hown in Fig. 7, th adaptiv compnation algorithm uing th wighting factor improv prformanc accuracy bttr than Ali algorithm and th uggtd compnation algorithm with no wighting factor. For xampl, in th ca of LEO atllit of altitud 1,000km, if th accuracy of th prdiction i rquird to b 99.5%, thn th adaptiv compnation algorithm tolrat th offt variation of th initial maurd Dopplr frquncy up to 60Hz. But undr th am condition, th prdiction algorithm with no wighting factor prmit th offt variation up to 105Hz, and Al prdiction algorithm prmit th offt variation up to 90Hz. In th ca of LEO atllit of altitud,000km, th adaptiv compnation algorithm with wighting factor provid narly prfct prdiction accuracy conidring th initial frquncy offt variation of up to 500Hz. 1 0 00 400 600 800 1000 100 1400 1600 Tim (c) Nw algoritm with wighting factor Nw algoritm w/o wighting factor Ali' algorithm Fig. 6. Comparion btwn th prdiction rror and th wighting factor. IV. SIMULATION RESULT 100 99.5 99 h=1,000km h=,000km Th prformanc of th two propod prdiction chm i imulatd, on without th wighting factor and th othr with th wighting factor. Thn, th imulation rult ar compard to tho of Ali chm. Th LEO atllit ytm i aumd to b on of two typ, on with a circular orbit of altitud 1000km or th othr of altitud,000km. Each atllit orbit i aumd to b inclind by 5 o. It i aumd that a ur trminal i locatd on th quator for th wort ca, and th initial input Dopplr frquncy offt vari in th rang of up to 500 Hz from th xact Dopplr frquncy for th practical ituation. In Fig. 7, w how th rult of th prdiction rror by imulation. W compar th accuracy of th prformanc btwn th prdiction algorithm, in trm of th cofficint of dtrmination factor, R dfind a following (0). Thi cofficint of dtrmination factor rprnt th normalizd accuracy. ( f D f D ) ( f D f DM ) ( f D f D ) R (0) = whr, f D i th xact Dopplr frquncy, f D i th avrag of f D, and f DM i th prdictd Dopplr frquncy by uing on of thr Dopplr prdiction algorithm mntiond abov, i.., Cofficint of Dtrmination (%) 98.5 98 97.5 97 96.5 96 95.5 95 0 100 00 300 400 500 Dopplr Input offt (Hz) Fig. 7. Prdiction accuracy comparion dpnding on dopplr input offt. Figur 8 how th cumulativ ditribution function of th prdiction rror for ach Dopplr frquncy prdiction algorithm. In th ca of LEO atllit of altitud 1,000km, th adaptiv compnation algorithm mt th prdiction rror rquirmnt which i gnrally rlatd with th rror tolranc lvl, for ETRI Journal, Volum, Numbr 4, Dcmbr 000 Moon-H You t al. 37
xampl 50Hz in typical GMPCS ytm. Howvr, othr algorithm can not atify thi rror rquirmnt. A for th propod prdiction algorithm without wighting factor, th probability i 17% in ca of th prdiction rror to b l than 50 Hz. A for Ali prdiction algorithm, th probability i only 0.7% in th am ca. Thi man that in ca of th prdiction rror to b mor than 50Hz th probability i 99.3% by uing Ali prdiction algorithm. In th ca of th adaptiv compnation algorithm with wighting factor in th atllit of altitud,000km, poibly occurrd prdiction rror i l than 50Hz. In Fig. 8 with concluion, th adaptiv compnation algorithm th mot applicabl mthod to b applid in th LEO atllit ytm of altitud from 1,000km to,000km for th gnral rror tolranc lvl, M=50Hz. CDF 1.0 0.8 0.6 0.4 0. Ali' algorithm Nw algorithm w/o wight Nw algorithm with wight 0.0 1 10 100 50 1000 10000 Prdiction rror (Hz) Fig. 8. Cumulativ ditribution function of th prdiction rror. V. CONCLUSION h=,000km h=1,000km In thi papr, w hav propod a Dopplr prdiction algorithm bad on th information of th intantanou atllit poition and th timatd trminal poition for LEO atllit ytm. In addition, nw adaptiv compnation algorithm for Dopplr frquncy hift wa uggtd from thi Dopplr prdiction algorithm with th wighting factor for th rduction of prdiction rror. For th prformanc valuation, th prdiction accuracy and th cumulativ dnity function of th prdiction rror for th propod algorithm wr imulatd and th rult wr compard to tho of Ali Dopplr prdiction chm. For th practical imulation rult, w conidrd th offt variation rang of th initially maurd Dopplr frquncy input and th rror tolranc lvl. From th imulation rult, th propod adaptiv compnation algorithm with wighting factor provid narly prfct prdiction accuracy conidring th initial frquncy offt variation of up to 500Hz in ca of LEO atllit of altitud,000km. And th adaptiv compnation algorithm i th mot applicabl mthod to b applid in th LEO atllit ytm of altitud from 1,000km to,000km for th gnral rror tolranc lvl, M= 50Hz. Th imulation rult how that our propod adaptiv algorithm i bttr than Ali algorithm. Although th propod chm nd a littl bit mor calculation load, th Dopplr prdiction rror rang that ha to b covrd by frquncy ynchronization circuit i much mallr. Thrfor, thi propod algorithm mak frquncy acquiition and tracking tp impl. REFERENCES [1] I. Ali, N. Al-Dhahir, and J. Hrhy, Dopplr Charactrization for LEO Satllit, IEEE Tran. Commun., Vol. 46, No. 3, Mar. 1998, pp.309 313. [] M. J. Millr, B. Vuctic, and B. L, Satllit Communication: Mobil and Fixd Srvic, Kluwr Acadmic Publihr, 1993. [3] S. Ohmori, H.Wakana, and S. Kawa, Mobil Satllit Communication, Artch Hou, 1998. [4] J. W. Jong, S. Sampi, and N. Morinaga, Larg Dopplr frquncy compnation tchniqu for trrtrial and LEO atllit dual mod DS/CDMA trminal, IEICE Tranaction on Communication, Vol. E79-B, No. 11, Nov. 1996, pp. 1,696 1,703. [5] E. Vilar and J. Autin, Analyi and Corrction Tchniqu of Dopplr Shift for Nongoynchronou Communication Satllit, Int. J. Satllit Commun., Vol. 9, 1991, pp.13 136. [6] A. Kajiwara, Mobil Satllit CDMA Sytm Robut to Dopplr Shift, IEEE Tran. Vh. Tchnol., Vol. 44, 1995, pp.480 486. [7] Y. Uno t al, Carrir Rgnaration in a Block Dmodulator for Low Earth-Orbital Satllit Communication Sytm, Proc. th 4th Int. Symp. Pronal, Indoor, and Mobil Radio Commun., 1993, pp.458 46. [8] M.H. You and S.I. L, Th Charactritic of Dopplr Shift ovr NGSO Satllit Communication Link, Proc. IEEK Summr Confrnc 98, Vol. 1, No. 1, Jun. 1998, pp. 6 9. [9] ITU-R, Rcommndation ITU-R M.15, App. to Annx, 1997. [10] Globaltar, Dcription of th Globaltar Sytm, Globaltar, L.P., Aug. 1995. [11] H. S. Mo t al, Dvlopmnt and Tting of Satllit Opration Sytm for Kora Multipurpo Satllit-I, ETRI Journal, Vol., No.1, March 000, pp.1 11. [1] C. Won t al, Miion Analyi and Planning Sytm for Kora Multipurpo Satllit-I, ETRI Journal, Vol. 1, No.3, Sp. 1999, pp.9 40. Moon-H You rcivd hr B.S. and M.S. in lctronic tlcommunication nginring from Hanyang Univrity, Soul, Kora, in 1988 and 1990, rpctivly. Sinc h joind ETRI in 1990, h ha bn involvd with dvlopmnt of atllit communication tchnology. Currntly, h i th had of Mobil Satllit Communica- 38 Moon-H You t al. ETRI Journal, Volum, Numbr 4, Dcmbr 000
tion Rarch Tam. Hr rarch intrt includ mobil atllit communication ytm and atllit communication application. Song-Pal L rcivd hi B.S. in lctrical nginring from Soul National Univrity in 1979. H rcivd th M.S. and Ph.D. dgr in lctrical nginring from Polytchnic Intitut of Nw York, USA, in 1986 and 1990, rpctivly. H joind ETRI in 1980, whr h i currntly working a a dirctor of Satllit Communication Sytm Dpartmnt, Radio and Broadcating Tchnology Laboratory. Hi rarch intrt includ Satllit Communication Sytm, Mobil Satllit Communication Srvic, Satllit Tracking Tlmtry & Command (TTC) Sytm and Stratophric Sytm. Youngyarl Han rcivd th B.S. dgr from Soul National Univrity, Soul, Kora, in 1960 and th M.S. and Ph.D. dgr from th Univrity of Miouri, Rolla, USA, in 1976 and 1979, rpctivly, all in lctrical nginring. Currntly, h i a Profor at Hanyang Univrity, Soul. Hi rarch intrt ar in th ara of tatitical ignal procing and communication thory a applid to analyi and valuation of digital communication ytm with a currnt mphai on mobil and wirl communication. H rcivd th National Ordr of Mrit and th Pridntial Citation from th Govrnmnt of Kora and th Achivmnt Award from th Kora Intitut of Communication Scinc (KICS). H wa th Vic Pridnt of KICS from 1990 to 1995. H i a Mmbr of Sigma Chi. ETRI Journal, Volum, Numbr 4, Dcmbr 000 Moon-H You t al. 39